Lightweight Electric/Hybrid Vehicle Design: Current EV design approaches (part 1)

Home | Articles | Forum | Glossary | Books

1.1 Introduction

The environmental arguments for electric propulsion become more compelling when they can be supported by an economic case that will appeal to the vehicle buyer. Here the current technology of electric and hybrid drive is reviewed in a way that shows the technical imperatives alongside the economic ones. After an analytical study of drive system comparisons for different vehicle categories, 'clean-sheet-design' integrated vehicle electric-drive systems are reviewed for small and medium cars and a concluding section encapsulates a procedure for optimizing motor, drive and batteries in the form of a power-pack solution.

A section on electric-drive fundamentals, establishing basic terminology, appears in the Introduction.

In the preface to the case study sections (5 and 6), contained in the second half of Section 4, the whole macro-economics of electric vehicles is discussed, with the wider aspects of the fuel infra-structure, as is a full analysis of competing electric-drive and energy-storage systems, for EVs.

1.2 Case for electric vehicles


ill. 1.1 Life expectancy related to energy usage, as seen by the World Bank. ENERGY USAGE

The current world population of motor vehicles stands at 700 million, of which over 600 million are owned in G7 economies 1 . This number is set to increase to around 1000 million in the next ten years. The bulk of this growth is expected to occur in Second World countries where per capita income is reaching levels where car ownership is known to commence. This has two serious implications (ill. 1.1): a large increase in the usage of hydrocarbon fuels and an increase in pollution to globally unsustainable levels. Much has been heard of the so-called Greenhouse Effect.

If carbon dioxide is on a scale of 1 as a greenhouse gas, methane is 25 and CFCs are 30 000- 50 000. Clearly the release of hydrocarbons and CFCs by man must be curtailed as soon as possible;

CO 2 is a different matter. If the quantity in the atmosphere was doubled from 20 to 40%, the temperature would increase by 5 o C and the sea level would rise by 1 meter. However, the additional plant activity would eliminate famine for millions in Africa , the Middle East and Asia . In scientific circles, the 'jury is still out' on carbon dioxide.

The problem emissions are those of carbon monoxide, sculpture dioxide, nitrous oxide and lead, not to mention solid particles from the exhausts of diesels. In all of these, man is competing with nature. The problem is that man's emissions are now set to reach levels which history shows have had dramatic consequences in nature. E.g., in 1815, a volcano emitted 200 million tons of sculpture dioxide into the atmosphere. In 18l6 there was a cloud of sulphuric acid in the sky which blocked out the sun in the northern hemisphere for the whole of the summer. The temperature fell by 7°C and there were no crops. Every 2000 megawatt power station which runs on coal emits 150 000 tons of sulphur dioxide per annum. Acid rain destroys our forests and buildings in the northern hemisphere. Pollution on this scale in the southern hemisphere is unsustainable. Nitrous Oxide is emitted when nitrogen burns at 1500° C or above. This gas reaches high concentrations in cities and is converted by sunlight into photosynthesis smog, which is becoming a major health hazard worldwide. A change in the technology of motor transport could have the fastest impact on this problem as most vehicles are replaced every ten years.


Consumers vote with their wallets! Electric vehicles will only have a healthy market based on a primary transport role using technology that achieves the performance of internal combustion engines. This means sources of energy other than batteries (ill. 1.2). In reality we have a choice of IC engine, gas turbine and fuel cell, but how can we maintain performance whilst reducing pollution? The secret is to stop wasting the 72% of energy that currently goes out of the exhaust pipe or up from the radiator. The IC engine is currently operated with a fuel/air ratio of 14:1. This can be increased to 34:1 but the engine can no longer accelerate rapidly. Fortunately, this can be overcome by other means. The gas turbine is an efficient solution for large engines over 100 kW in commercial vehicles. Its performance isn't as good as an IC engine's at lower powers, however, and fuel-cell electrics offer the best promise. Fuel cells are the technology of the future. There are many sorts but only one type of any immediate relevance to vehicles and this is the proton exchange membrane (PEM) cell. Using the Carnot cycle, this has a conversion efficiency limit of 83%.

Scientists can achieve 58% now and are predicting 70% within ten years. Fuel cells have many excellent qualities. Small units are efficient - especially at light load. New construction techniques are reducing costs all the time and £200/kW was already achievable in 1992 using a hydrogen/air mixture. The real problem is providing the fuel.


Current engines obtain their energy by burning hydrocarbons such as propane, methane, petrol, diesel and so on. However, hydrogen is the fuel of the future. What powers a Saturn 5 Moon Rocket? Coincidence, or sheer necessity? Liquid hydrogen has an energy density of 55 000 BTUs per pound compared to 19 000 BTUs per pound for petrol and 17 000 BTUs per pound for propane. The problem is obtaining large amounts of hydrogen efficiently from hydrocarbon fuels. The percentage of hydrogen directly contained in these fuels is small in energy terms. E.g., methane (CH 4 ) has 17.5% of its energy in carbon and 25% in hydrogen. However, l. Petrol car: A journey of 68 miles each day consumes 2.5 gallons of fuel and takes 2 hours.

Amount of energy in fuel = 5.14 x 10 8 joules Thermal power = 71.3 kW Mechanical power = 20 kW average Efficiency = 28%

2. Battery electric car as secondary transport.

Power station efficiency 40% Electric car efficiency 80% OVERALL 32% CONCLUSION: Pollution is moved from car to power station. There is only an environmental return if the car’s performance is sacrificed or the power station is non-thermal and range/ performance is limited.

3. Hybrid car as primary transport.

Hydrocarbon to electricity Via lean burn petrol engine 45% Electricity to mechanical power 90% OVERALL 40.5% CONCLUSION: Pollution reduced by 55% and fuel consumption is 70% of petrol vehicle with performance/range as the petrol vehicle.

4. Fuel-cell electric car as primary transport.

Hydrocarbon to hydrogen conversion 80% Fuel-cell hydrogen to electricity 60% Electricity to mechanical power 90% OVERALL 43% (potential for 48% in 10 years) CONCLUSION: Pollution reduced by 90%; fuel consumption is 66% of petrol vehicle and performance/range is as petrol vehicle.

ill. 1. 2 Some crude comparisons for fuel related to pollution.

there is now a solution to this problem, with a reforming process developed by Hydrogen Power Corporation/Engelhard called Thermal Catalytic Reforming. Put simply, it's the chemical process:

3Fe + 4H 2 O = Fe 3 O 4 + 4H 2 and Fe 3 O 4 + 2C = 3Fe + 2CO 2

The first process takes place with a catalyst at 130° C. The hydrogen is stored in a hydride tank until required. The iron is returned to a central facility for reduction by the second process. The main points about this cycle are that a high proportion of hydrocarbon heat energy is converted into hydrogen and that 1 kg of iron provides enough hydrogen for a small car to travel 6 km on a fuel cell.

IC engines and gas turbines run well on most hydrocarbons and hydrogen. Fuel cells need hydrogen. Hydrogen has to be used and stored safely. This could be achieved by reforming it on demand at fuel stations - the waste heat would be used to generate electricity to be pumped back into the national grid. The primary fuel could be any hydrocarbon such as petrol, diesel, methanol, propane or methane. The only constraint is that the fuel source must have low sulphur content so as not to poison the catalyst. In the UK , we have a head start called the Natural Gas Grid. This is likely to become of critical importance for energy distribution, removing the need to distribute petrol and diesel by road. To satisfy future transport needs, we retain our 'fuel' stations as the means of distribution. This brings us to the problem of on-board hydrogen storage.

Iron titanium hydride has long been known as a storage medium but one would need 500 kg to store 10 liters of hydrogen, at a cost of £3000 in 1992. The gas is stored in a standard propane tank filled with this material. If the tank is ruptured, the gas is given off slowly because of its absorption in the hydride. In the USA experiments are also taking place with cryogenic storage which is potentially cheaper and lighter. The overall distribution scheme is illustrated in ill. 1.3. To summarize, the benefits of a change to hybrid/fuel-cell electric vehicles are: (i) engineering is practical; (ii) performance is acceptable to the consumer; (iii) it reduces fuel consumption; (iv) it reduces pollution, especially Nitrous oxide; (v) it reduces dependence on imported oil; (vi) it can be achieved quickly; (vii) it can be achieved at sensible cost; (viii) it prevents increased demand for oil; (ix) it fits in with the existing fuel infrastructure and (x) it solves the pollution problem in relation to projected pollution levels, not existing ones - the prime cause of the catalytic converter being ineffective.


This vehicle category, ill. 1.4, will use a fuel cell to provide the motive power for the average power requirement and utilize a booster battery to provide the peak power for acceleration. Hydrogen would be stored in a tank full of metal hydride powder, or cryogenically. This system provides enough waste heat for cabin heating purposes. The fuel cell can recharge the battery when the vehicle isn't in use. If the vehicle has an AC drive, it's possible for it to generate electricity for supply to portable tools, a house, or injection into the national grid. Fuel cells should reduce emission levels by a factor of 10, compared with IC engines on 14:1 air: fuel mixture.


What is a fuel cell? It is an electrochemical cell which converts fuel gas and oxidant into electricity and water plus waste heat (see Section 4). The PEM cell has graphite electrodes with a layer of membrane sandwiched in between, plus gas-tight seals. Each cell is about 6 mm thick and produces 1 V off-load and 0.7 V on-load, at a current of around 250 amps. Consequently a fuel cell for a 15

ill. 1.3 Hydrogen distribution system.

ill. 1.4 Fuel-cell electric vehicle.

kW average power would produce about 60-70 V DC at 250 amps. In size it would be about 200 mm square and about 600 mm long. The cell operates at a temperature of 80 ° C. When cold, it can give 50% power instantly and full power after about 3 minutes. The units exhibit very long life.

The problem until recently has been seal life when operated on air as opposed to oxygen. New materials have solved this problem. Output doubles when pure oxygen is used. Fuel cells don't like pollutants such as carbon monoxide in the source gases. Gas is normally injected at 0.66 atmospheres into the stack. The main challenge now is to refine the design so as to optimize the cost relative to performance. This will take time because the effort deployed at this time is small in relation to the effort put into batteries or other fuel-cell types. There is a very real case for a major multinational effort to train scientists and engineers in this technology in the short term, and to reduce the time to introduction on a large scale.


Batteries have been with us for at least 150 years and have two main problems: they are heavy and they don't like repeated deep discharge. Batteries which are deep cycled, irrespective of the technology, deteriorate in performance with age. So the question must be asked 'what can batteries do well?'. The answer is to provide limited performance in deep discharge, or alternatively, much better performance as a provider of peak power for hybrid and fuel-cell vehicles.

Much work is under way on high temperature cells. These are unlikely to meet cost or weight constraints of primary transport applications. The best high temperature batteries can offer 100 Wh/kg. Overall, fuel cells already give 300 Wh/kg and this can be improved with development.

What is needed is a battery with different capabilities to normal car starter batteries, namely very low internal resistance, long life, excellent gas recombination, room temperature operation, totally sealed, compact construction, reasonable deep discharge life as well as being physically robust.

The battery which satisfies the above criteria is the lead-acid foil battery, as manufactured by Hawker Siddeley. This type of construction has replaced nickel-cadmium pocket batteries on many aircraft. In particular the lead-acid foil battery retains far more charge from regeneration than conventional designs and can be charged and discharged rapidly. However, there is a trick to achieving this. Most batteries are made up of 'rectangular' arrays of cells so it's no wonder that the temperature of the cells varies with position in the stack. To charge a battery quickly it's vital to keep the cells at an even temperature. Consequently it's necessary to liquid cool the cells so as to obtain best performance and long life. Other points worthy of note are that batteries work best when hot; 40°C is ideal for lead-acid. The battery electrolyte is just the place to dump waste heat from the motor/engine/fuel cell.

Nickel-cadmium batteries offer better performance than lead-acid but are double the cost per Wh of storage at present and sealed versions are limited to 10 Ah but larger units are under development. The best nickel-cadmium units available at present are the SAFT STM/STH series.

Sealed lead-acid and aqueous nickel-cadmium cells have peak power in W/kg of 90 and 180, with Wh/kg values being 35 and 55 respectively.

In terms of safety, long series strings of aqueous batteries are not a good idea. The leakage from tracking is high and they are very dangerous to work on. Consequently batteries should be of sealed construction with no more than 110 V in a single string. Ideally, the maximum voltage should be 220 V DC, that's +/ - 110 V to ground arranged as two separate strings with a centre tap, so that no more than 110 V appears on a connector, with respect to ground, ill. 1.5.

ill. 1.5 Battery connections and earthing. There is an opening in the market place for a low cost 2 pole, 220 V, 300 A remote-control circuit breaker to act as battery isolator with 5 kA short-circuit capacity. However, there is a problem with earthing the centre tap of the battery as one may need an isolating transformer in the battery charger. Consequently, in many of the new schemes proposed in the USA , a different route is implemented which is used in trolley buses - the all-insulated system. In most of these schemes, large capacity batteries are used (15-30 kWh) at a typical nominal voltage of 300 V. This will vary from 250 V fully discharged to 375 V at the end of charging.

The electrical system is fully insulated from earth. During charging, the mains supply can be either centre tap ground or one-end ground. In the centre tap ground (typical USA situation, with 110/0/110) the potential of the vehicle electrics is balanced to earth. When one end is earth (typical European situation) the potential of the vehicle electrics will move up and down at the supply frequency with respect to ground and there is the prospect of earth leakage current through any capacitance to earth of the vehicle electrical system. However, this is very small, usually because the tyres isolate the vehicle. However, when charging it would be desirable to ground the vehicle body to prevent any shocks from people touching the vehicle and standing on a grounded surface.


From the previous considerations one can now start the task of specifying EV capability/ performance trade-offs. Polaron believe EVs will be partitioned as shown in ill. 1.6. This does not pretend to be an exhaustive list but to show the range and scale of requirements to be provided for. The most interesting observation is that in the mass market, 30-150 kW, a solution is possible using just two sizes of drive, 45 and 75 kW. To complement the drives, motors are required of two speed ratings for each size, say 5000 rpm where compatibility with a prime mover is required, and 12 000 rpm for the direct drive series hybrid/pure electric case.



Power GVW Engine Motor; Motor Turbo Application Rating type rating alternator

Below Less than None Brush DC Up to None Straight battery 40 kW 2 tons 40 KW electric van or car 40 kW 2 ton IC Brushless DC 1 x 45 kW None Parallel hybrid

-150 kW 5000 rpm family car 2 ton GT Brushless DC 1 x 75 kW 1 x 100 kW Parallel hybrid 12€000 rpm 60€000 rpm performance saloon 3 ton IC Brushless DC 1 x 75 kW None Parallel hybrid 5000 rpm 1 ton truck 5 ton GT Brushless DC 2 x 45 kW 1 x 100 kW Series hybrid 12€000 rpm 60€000 rpm 2 ton truck 7 ton GT Brushless DC 2 x 75 kW 1 x 150 KW Series hybrid 12€000 rpm 50€000 rpm single deck bus 150 kW 10 ton GT Switched 1 x rating 1 x rating Heavy traction to reluctance 5000 rpm 50€000 rpm and road haulage motor at 150 kW 1 MW 25€000 rpm Series hybrid 40 tons at 1 MW configuration


ill. 1.6 Short-term battery electric and hybrid vehicles.

It is now proposed to have a look at two cases (a) 45 kW parallel hybrid vehicle; (b) 90 kW series hybrid vehicle, as in (ill. 1.7). The 45 kW parallel hybrid vehicle consists of, typically, a small engine driving through a motor directly into the differential gear and hence to the road wheels. Minimization of weight is the key issue on such a design along with low rolling resistance and low drag. At 60 mph a good design can expect to draw 8 kW to keep going on a flat level road. The vehicle would be fitted with an engine rated to supply about one-third of the peak requirement, that's 15 kW plus an allowance for air conditioning if relevant. The motor has to deliver up to 45 kW using energy stored in batteries. This can be done either by a constant torque motor operating via a gearbox or a constant power motor operating with only two gears or without a gearbox. The latter is rapidly becoming the standard for EVs using front wheel drive.

The vehicle uses the battery to provide peak acceleration power for overtaking, hill climbing and so on. On the flat a 0-60 mph acceleration time of around 12 seconds would be typical for this class of vehicle and a top speed of perhaps 80 mph, where permitted; the engine is started when the road speed exceeds 20 mph and then clutched into the motor. The engine then charges the batteries as well as satisfying the average demand of the car. During acceleration the electric drive and the engine work together to provide peak acceleration. It is in the cruise condition that optimum efficiency is required. Consequently more sophisticated designs use 3 way clutch units so that the motor can be mechanically disconnected when the battery is fully charged and only switched back in for acceleration. In this condition attention must also be paid to the minimization of rolling resistance and windage losses (Figs 1.6 and 1.8(a)).

The series hybrid vehicle corresponds to a high performance sports saloon. A 0-60 mph time of 7 seconds and a top speed of 120 mph could be expected (Figs 1.7 and 1.8(b)). The main power source would be a gas turbine which would operate through a PWM inverter stage to feed 300- 500 V DC into the main bus. There are two separate drives, each driving a rear wheel of the vehicle. To reduce weight, the motors would be designed for 12 000 rpm and gearboxes employed to reduce the speed to the road wheels - about 1800 rpm at 120 mph. The gas turbine may operate over a 2:1 speed range to give good efficiency. Specific fuel consumption is doubled at 15 kW compared to 100 kW. However, overall consumption would still be that of a 'Mini', with emissions to match. The peak power for acceleration would come from batteries - probably nickel-cadmium in this case, where cost pressures are not so demanding.

ill. 1.7 A 45 kW parallel hybrid and 90 kW series hybrid.

ill. 1.8 Torque-speed curves for 45 kW vehicle (a) and each motor.

ill. 1.9 Fuel-cell power conversion.


What are the design problems for the electrical system? The first one is cost. Unless the final product is attractive to the consumer, we don't have a market. Where are we now? For 1000 off systems at 45 kW, a brushless DC motor would cost £1000, a controller £2000, and a battery £2000 (lead-acid). These 1992 prices will reduce with mass production. The second design challenge is one of methodology. Electric vehicles have been traditionally built by placing motor and batteries then spreading the electrical system over the vehicle. This needs to change. Polaron would like to suggest a modular approach to the problem whereby sealed batteries and controller power electronics are in one unit and the motor is in fact the second. The third design challenge is one of compatibility. Low performance vehicles can be built with 110 V electrical systems. However, as the power increases this isn't practical. But both fuel cells and batteries are low voltage heavy current devices - how can this conflict be addressed? The solution is to use power conversion. In ill. 1.9 a 100 stage fuel cell is integrated with a 216 V battery to give a stabilized 300 V DC rail. The motor and controller are then built at 300 V where the currents are significantly reduced on the 100 V system. As the power level rises, voltages up to 500 V DC can be anticipated. However, when the power conversion is switched off the highest voltage will be the battery voltage. This additional power conversion will be needed for another reason. If vehicles are equipped with small booster batteries for acceleration, the DC link voltage will change significantly according to load conditions. The power conversion provides a means of stabilizing for this variation.

1.2.10 MOTOR TYPES AND LOCATIONS (ill. 1.10)

Which is the best type of motor? Answer - the cheapest. Which is the cheapest motor? Answer - the lightest. Which is the lightest motor? Answer - the most efficient. On this criteria, there is no doubt that a permanent magnet brushless DC motor would sweep the board. However, our enthusiasm must be tempered by two other considerations, cost of materials and controller costs.

The factors affecting selection are covered in Section 1.3.


ill 1.11 illustrates a typical pure battery electric vehicle scheme which could also be used in hybrid mode with an engine if required. The motor is a shunt field unit such as the Nelco Nexus 2 unit used in many industrial EVs. This machine is a 4 pole motor with interpoles and operates at a maximum voltage of 200 V DC. The field supply is typically 30 amps for maximum torque.

The controller consists of a 2 quadrant chopper with a switch capacity of 400 amps. An electromechanical contactor shorts out the positive chopper switch in cruise mode for maximum efficiency. The chopper is fitted with input RF filtering and pre-charge to extend contactor life.

The chopper switches at 16 kHz and the output contains a small L/C filter to remove the dv/dt from the machine armature. A Hall effect DCCT measures the armature current for the control system.

ill. 1.10 Motor specifications. Nelco electric 34 kW brush motor specification

In the power supply area, there are four components: first is the battery charger, in this case a CUK converter, or a boost/buck chopper is also a possibility to make the mains current look like a sine wave for ensuring IEC555 compliance. Control of battery charging conditions is one of the most important considerations in extending battery life in deep discharge. For lead-acid batteries the level of float voltage is critical as well as maintaining cell temperature. The battery charger could incorporate a 20 kHz isolating transformer if costs permit. Experiments are under way with inductive power transfer which isolates the car and makes it necessary to plug in for charging. Another possibility is an automatic self-aligning connector which the car drives into when parking. The next consideration is the auxiliary 13.6 V battery supply. The vehicle seems likely to retain a separate 12 V battery for lighting and control functions. A 300 W DC/DC converter will satisfy this requirement. The third consideration is the control system power.

This is a small (20 W) DC/DC converter which provides the control power for the chopper. It is likely to be incorporated with the main control PCB and could also be supplied from the 13.6 V battery. The final factor is the field controller. This is a 4 quadrant chopper which provides the motor field supply. It has to be able to reverse the current so that the motor can reverse without contactors in the armature circuit. If the motor has a tachometer fitted, this may be used for braking control and blending with electromechanical brakes. The important issue with this controller is that the power switching is contained in a single unit so that all the DC components are kept in one place. This is important for another reason to meet IEC555 RF interference legislation. Therefore all insulated systems will require an isolated conductive casing which can be connected to vehicle chassis.


This is illustrated in ill. 1.12. The drive consists of a 3 phase PWM Drive which feeds the 3 phase motor. The beauty of this arrangement is that the motor can be disconnected and the mains fed to the inverter arms to give a high power battery charger, by phase locking the PWM to the mains.

ill. 1.11 Controller for 34 kW shunt field DC motor.

An alternative to this arrangement is for the inverter to put power back into the mains. In case of fault, three alternistors provide current limit protection. In the brushless DC case, the motor permanent magnets provide 50% of the flux and the remainder comes from a 50 amp circulating current Id at right angles to the torque producing component Iq.

The inverter is constructed using 300 amp IBGT phase leg packages which minimize the inductance between transistors and associated bypass diodes. The inverter output is filtered by 6 x 10 mH capacitors plus 3 x 5 mH inductors. This reduces the 18 kHz carrier ripple current in the motor to about 20 AP/P. There is a real time digital signal processor (DSP) which performs vector control using state space techniques and this includes 3rd harmonic injection to maximize the inverter output voltage. Comprehensive overload protection is fitted. The inverter demand is a torque signal and a speed feedback is provided for the vehicle builder to close the speed loop.

Both signals are PWM format (10-90%) on a 400 Hz carrier. The drive can be adapted for induction motor control but this isn't so efficient, as explained in the motor section below.


ill 1.13 illustrates a turbo alternator scheme for gas turbines. This scheme has two purposes: it starts the turbine, and provides a stabilized DC link voltage for a 2:1 change in turbine speed and changes in DC link current from no-load to full-load. The alternator itself is the result of many years' development in high speed gas compressors. It is a 4 pole unit which allows iron losses to be kept low and in particular the tooth tip temperature reasonable whilst still using silicon steel laminations (2 pole permanent magnet alternators are potential fireballs!). The magnet material is samarium cobalt with a carbon fiber or Kevlar sleeve. At these speeds, one needs every bit of strength possible. The magnets are capable of operation at 150° C. The use of metallic magnets isn't a problem here because the weight is small. Hall sensors are fitted for machine timing during starting and voltage control purposes. A small L/C filter limits the amplitude of the carrier ripple on the alternator windings.

ill. 1.12 Electric vehicle 45 kW inverter.


From the foregoing considerations, it will be apparent that the motor car of the future needs power electronics to be viable. Fortunately, we now have the technology to satisfy the most demanding applications. There may be some rivalry between different types of power switches but cost will be the final judge. A manufacturer who constructs the power electronics as an all-insulated system in a single module permits module exchange as the first means of maintenance. Liquid cooling also makes sense. It can cool the motor, warm/cool the sealed batteries and provide power steering at the same time. This concept will make it possible to convert existing chassis as well as develop new ones, thus enabling product to be brought to market quickly. Standard electronics packages are the only way to achieve the unit costs necessary for product acceptance in the market.

Interchangeable batteries will make it possible for maximum vehicle utilization in intensive duty applications, such as taxis and delivery vehicles. This method of construction also opens the door to new methods of financing EVs; for example, the user buys vehicle then rents battery/power electronics.

cont. >>

Top of Page